rRNADepletion_s4ter

Setup

library(tidyverse)
library(magrittr)
library(ngsReports)
library(here)
library(scales)
library(ggpubr)
library(kableExtra)
library(AnnotationHub)
library(ensembldb)
library(edgeR)
library(corrplot)
library(DT)
library(pander)
library(ggrepel)
library(pheatmap)
library(ggdendro)
library(Biostrings)
library(Gviz)
library(Rsamtools)
library(viridis)
library(ggfortify)

if (interactive()) setwd(here::here())
theme_set(theme_bw())
cores <- 6

Sequence information

ah <- AnnotationHub() %>%
  subset(species == "Homo sapiens") %>%
  subset(rdataclass == "EnsDb")
ensDb <- ah[["AH83216"]]
trEns <- transcripts(ensDb) %>%
  mcols() %>% 
  as_tibble()
trLen <- exonsBy(ensDb, "tx") %>%
  width() %>%
  vapply(sum, integer(1))
geneGcLen <- trLen %>%
  enframe() %>%
  set_colnames(c("tx_id", "length")) %>%
  left_join(trEns) %>%
  group_by(gene_id) %>% 
  summarise(
    aveLen = mean(length),
    maxLen = max(length), 
    aveGc = sum(length * gc_content) / sum(length),
    longestGc = gc_content[which.max(length)[[1]]]
  ) %>%
  mutate(
    aveGc =  aveGc / 100,
    longestGc = longestGc / 100
  )
trGcLen <- trLen %>%
  enframe() %>%
  set_colnames(c("tx_id", "length")) %>%
  left_join(trEns) %>%
  group_by(tx_id) %>% 
  summarise(
    aveLen = mean(length),
    maxLen = max(length), 
    aveGc = sum(length * gc_content) / sum(length),
    longestGc = gc_content[which.max(length)[[1]]]
  ) %>%
  mutate(
    aveGc =  aveGc / 100,
    longestGc = longestGc / 100
  )
genesGR <- genes(ensDb)
mcols(genesGR) <- mcols(genesGR)[c("gene_id", "gene_name", 
                                         "gene_biotype", "entrezid")]
txGR <- transcripts(ensDb)
mcols(txGR) <- mcols(txGR)[c("tx_id", "tx_name", 
                                   "tx_biotype", "tx_id_version", "gene_id")]

An EnsDb object was obtained for Ensembl release 101 using the AnnotationHub package. This provided the GC content and length for every gene and transcript in the release. For humans, this consists of 67990 genes and 252335 transcripts.

Raw data

Sample information

files <- list.files(
  path = "/hpcfs/users/a1647910/20200310_rRNADepletion/4_T47D_ZR75_DHT_StrippedSerum/0_rawData/FastQC",
  pattern = "zip",
  full.names = TRUE
)

samples <- tibble(
  sample = str_remove(basename(files), "_fastqc.zip"),
  dataset = NA,
  organism = NA
) %>%
  mutate(
    dataset = "StrippedSerum",
    organism = "human"
  )
datasets <- samples$dataset %>% 
  unique()

The following analysis involves 16 paired-end samples across 1 dataset(s): StrippedSerum.

Library sizes

rawFqc <- files %>%
  FastqcDataList()

data <- grep("GLL", fqName(rawFqc))
labels <- rawFqc[data] %>%
  fqName() %>%
  str_remove("_001\\.fastq\\.gz") %>%
  str_remove("Ps2Ex3M1_")
rawLib <- plotReadTotals(rawFqc[data]) +
  labs(subtitle = "StrippedSerum") + 
  scale_x_discrete(labels = labels)

The library sizes of the unprocessed dataset(s) range between 22,057,776 and 42,743,934 reads.

rawLib

GC content

rRNA transcripts are known to have high GC content. Therefore, inspecting the GC content of the raw reads is a logical start point for detecting incomplete rRNA removal. A spike in GC content at ~ 70% is expected if this is the case.

plotly::ggplotly(
  plotGcContent(
    x = rawFqc[data], 
    plotType = "line",
    gcType = "Transcriptome",
    species = "Hsapiens"
  ) +
    labs(title = "StrippedSerum Dataset (H. sapiens)") + 
    theme(legend.position="none")
) 

GC content of reads in the dataset. Clear spikes at about 70% GC are observed, which is likely due to incomplete rRNA depletion.

Overrepresented sequences

getModule(rawFqc, "Overrep") %>% 
  group_by(Sequence, Possible_Source) %>% 
  summarise(`Found In` = n(), `Highest Percentage` = max(Percentage)) %>% 
  arrange(desc(`Highest Percentage`), desc(`Found In`)) %>% 
  ungroup() %>% 
  dplyr::slice(1:30) %>%
  mutate(`Highest Percentage` = percent_format(0.01)(`Highest Percentage`/100)) %>%
  kable(
    align = "llrr", 
    caption = paste(
      "Top", nrow(.),"Overrepresented sequences.",
      "The number of samples they were found in is shown,",
      "along with the percentage of the most 'contaminated' sample."
    )
  ) %>%
  kable_styling(
    bootstrap_options = c("striped", "hover", "condensed", "responsive")
  )

Top 30 Overrepresented sequences. The number of samples they were found in is shown, along with the percentage of the most ‘contaminated’ sample.

Sequence	Possible_Source	Found In	Highest Percentage
CCTTAGGCAACCTGGTGGTCCCCCGCTCCCGGGAGGTCACCATATTGATG	No Hit	16	0.84%
CTGGAGTCTTGGAAGCTTGACTACCCTACGTTCTCCTACAAATGGACCTT	No Hit	16	0.60%
CTCCTTAGGCAACCTGGTGGTCCCCCGCTCCCGGGAGGTCACCATATTGA	No Hit	16	0.59%
CCCAAACCCACTCCACCTTACTACCAGACAACCTTAGCCAAACCATTTAC	No Hit	8	0.55%
CCCCTCCTTAGGCAACCTGGTGGTCCCCCGCTCCCGGGAGGTCACCATAT	No Hit	16	0.50%
CCAGGCTGGAGTGCAGTGGCTATTCACAGGCGCGATCCCACTACTGATCA	No Hit	13	0.50%
CCCTCCTTAGGCAACCTGGTGGTCCCCCGCTCCCGGGAGGTCACCATATT	No Hit	16	0.49%
CTTGGTTATAATTTTTCATCTTTCCCTTGCGGTACTATATCTATTGCGCC	No Hit	16	0.49%
CTCCGTTTCCGACCTGGGCCGGTTCACCCCTCCTTAGGCAACCTGGTGGT	No Hit	16	0.47%
CTTAGGCAACCTGGTGGTCCCCCGCTCCCGGGAGGTCACCATATTGATGC	No Hit	15	0.45%
CCCTGTTCTTGGGTGGGTGTGGGTATAATGCTAAGTTGAGATGATATCAT	No Hit	8	0.41%
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA	No Hit	1	0.40%
CCCAAACCCACTCCACCCTACTACCAGACAACCTTAGCCAAACCATTTAC	No Hit	8	0.39%
CCCTGTTCTTGGGTGGGTGTGGGTATAATACTAAGTTGAGATGATATCAT	No Hit	8	0.39%
GTATAATACTAAGTTGAGATGATATCATTTACGGGGGAAGGCGCTTTGTG	No Hit	8	0.39%
CGGGGGAAGGCGCTTTGTGAAGTAGGCCTTATTTCTCTTGTCCTTTCGTA	No Hit	16	0.36%
GTGGGTATAATGCTAAGTTGAGATGATATCATTTACGGGGGAAGGCGCTT	No Hit	8	0.33%
CTCTCTACAAGGTTTTTTCCTAGTGTCCAAAGAGCTGTTCCTCTTTGGAC	No Hit	16	0.33%
CTTATTTCTCTTGTCCTTTCGTACAGGGAGGAATTTGAAGTAGATAGAAA	No Hit	16	0.32%
GGGTATAATACTAAGTTGAGATGATATCATTTACGGGGGAAGGCGCTTTG	No Hit	8	0.31%
CTCAGACCGCGTTCTCTCCCTCTCACTCCCCAATACGGAGAGAAGAACGA	No Hit	12	0.31%
GTAAGATTTGCCGAGTTCCTTTTACTTTTTTTAACCTTTCCTTATGGGCA	No Hit	8	0.31%
CTGAACTCCTCACACCCAATTGGACCAATCTATCACCCTATAGAAGAACT	No Hit	16	0.29%
GTCCAATTGGGTGTGAGGAGTTCAGTTATATGTTTGGGATTTTTTAGGTA	No Hit	12	0.28%
CCTCCTTAGGCAACCTGGTGGTCCCCCGCTCCCGGGAGGTCACCATATTG	No Hit	16	0.28%
CTGGTTTCGGGGGTCTTAGCTTTGGCTCTCCTTGCAAAGTTATTTCTAGT	No Hit	15	0.28%
CTCTAGAATAGGATTGCGCTGTTATCCCTAGGGTAACTTGTTCCGTTGGT	No Hit	16	0.26%
CTGTTCTTGGGTGGGTGTGGGTATAATACTAAGTTGAGATGATATCATTT	No Hit	5	0.25%
CTCCGAGGTCGCCCCAACCGAAATTTTTAATGCAGGTTTGGTAGTTTAGG	No Hit	16	0.25%
CTCAGGCTGGAGTGCAGTGGCTATTCACAGGCGCGATCCCACTACTGATC	No Hit	13	0.25%

Trimmed data

Raw libraries were trimmed using cutadapt v1.14 to remove Illumina adapter sequences. Bases with PHRED score < 30, NextSeq-induced polyG runs and reads shorter than 35bp were also removed.

trimFqc <- list.files(
  path = "/hpcfs/users/a1647910/20200310_rRNADepletion/4_T47D_ZR75_DHT_StrippedSerum/1_trimmedData/FastQC",
  pattern = "zip",
  full.names = TRUE
) %>%
  FastqcDataList()
trimStats <- readTotals(rawFqc) %>%
  dplyr::rename(Raw = Total_Sequences) %>%
  left_join(readTotals(trimFqc), by = "Filename") %>%
  dplyr::rename(Trimmed = Total_Sequences) %>%
  mutate(
    Discarded = 1 - Trimmed/Raw,
    Retained = Trimmed / Raw
  )

After trimming of adapters between 6.09% and 8.36% of reads were discarded.

Aligned data

Trimmed reads were:

1. Aligned to rRNA sequences using the BWA-MEM algorithm to estimate the proportion of reads that were of rRNA origin within each sample. BWA-MEM is recommended for high-quality queries of reads ranging from 70bp to 1Mbp as it is faster and more accurate that alternative algorithms BWA-backtrack and BWA-SW.

2. Aligned to the Homo sapiens GRCh38 genome (Ensembl release 101) using STAR v2.7.0d and summarised with featureCounts from the Subread v1.5.2 package. These counts were used for all gene-level analysis.

rRNA proportions

rRnaProp <- read.delim(
  "/hpcfs/users/a1647910/20200310_rRNADepletion/4_T47D_ZR75_DHT_StrippedSerum/3_bwa/log/samples.mapped.all", 
  sep = ":", 
  col.names = c("sample", "proportion"), 
  header = FALSE
) %>% 
  mutate(
    sample = str_remove_all(sample, ".mapped"),
    sample = basename(sample),
    proportion = proportion/100,
    dataset = "StrippedSerum",
    organism = "human"
  ) %>%
  as_tibble()
rRnaProp$dataset %<>%
  factor(levels = c("StrippedSerum"))

rRnaProp %>%
  ggplot(aes(sample, proportion)) +
  geom_bar(stat = "identity", position = "dodge") +
  facet_wrap(~dataset, scales = "free_x") +
  scale_y_continuous(labels = percent) +
  labs(x = "Sample", y = "Percent of Total", fill = "Read pair") +
  theme(axis.text.x = element_text(angle = 90, hjust = 1))

Percentages of each library that align to rRNA sequences with bwa mem.

Gene GC content and length

dgeList <- read_tsv("/hpcfs/users/a1647910/20200310_rRNADepletion/4_T47D_ZR75_DHT_StrippedSerum/4_star2pass/featureCounts/genes.out") %>%
  set_colnames(basename(colnames(.))) %>%
  set_colnames(str_remove(colnames(.), "Aligned.sortedByCoord.out.bam")) %>%
  as.data.frame() %>%
  column_to_rownames("Geneid") %>%
  DGEList() %>%
  calcNormFactors()
metaData <- read_tsv("/hpcfs/users/a1647910/20200310_rRNADepletion/4_T47D_ZR75_DHT_StrippedSerum/T47D_ZR75_DHT_StrippedSerum.tsv")

dgeList$genes <- genesGR[rownames(dgeList),]
mcols(dgeList$genes) %<>% 
  as.data.frame() %>% 
  left_join(geneGcLen)
addInfo <- tibble(
  sample = rRnaProp$sample,
  dataset = "StrippedSerum",
  organism = "human",
  rRNA = rRnaProp$proportion
) %>%
  left_join(metaData)
dgeList$samples %<>%
  rownames_to_column("rowname") %>%
  mutate(sample = rowname) %>%
  left_join(addInfo) %>%
  column_to_rownames("rowname")
dgeList$samples$filenames <- list.files(
  "/hpcfs/users/a1647910/20200310_rRNADepletion/4_T47D_ZR75_DHT_StrippedSerum/2_alignedData/bam", 
  pattern = ".bam$", 
  full.names = TRUE
)
# dgeList$samples$group <- colnames(dgeList) %>%
#   str_extract("(D|V)") %>%
#   factor(levels = c("D", "V"))

gcInfo <- function(x) {
  x$counts %>%
    as.data.frame() %>%
    rownames_to_column("gene_id") %>%
    as_tibble() %>%
    pivot_longer(
      cols = colnames(x), 
      names_to = "sample", 
      values_to = "counts"
    ) %>%
    dplyr::filter(
      counts > 0
    ) %>%
    left_join(
      geneGcLen
    ) %>%
    dplyr::select(
      ends_with("id"), sample, counts, aveGc, maxLen
    ) %>%
    split(f = .$sample) %>%
    lapply(
      function(x){
        DataFrame(
          gc = Rle(x$aveGc, x$counts),
          logLen = Rle(log10(x$maxLen), x$counts)
        )
      }
    ) 
}
gcSummary <- function(x) {
  x %>%
    vapply(function(x){
      c(mean(x$gc), sd(x$gc), mean(x$logLen), sd(x$logLen))
    }, numeric(4)
    ) %>%
    t() %>%
    set_colnames(
      c("mn_gc", "sd_gc", "mn_logLen", "sd_logLen")
    ) %>%
    as.data.frame() %>%
    rownames_to_column("sample") %>%
    as_tibble()
}

rle <- gcInfo(dgeList)

sumGc <- gcSummary(rle)

a <- sumGc %>%
  left_join(dgeList$samples) %>%
  ggplot(aes(rRNA, mn_logLen)) +
  geom_point(aes(colour = treat, shape = cell_line), size = 3) +
  geom_smooth(method = "lm") +
  scale_x_continuous(labels = percent) +
  labs(
    x = "rRNA Proportion of Initial Library",
    y = "Mean log(Length)",
    colour = "Treatment",
    shape = "Cell line"
  ) 
b <- sumGc %>%
  left_join(dgeList$samples) %>%
  ggplot(aes(rRNA, mn_gc)) +
  geom_point(aes(colour = treat, shape = cell_line), size = 3) +
  geom_smooth(method = "lm") +
  scale_y_continuous(labels = percent) +
  scale_x_continuous(labels = percent) +
  labs(
    x = "rRNA Proportion of Initial Library",
    y = "Mean GC Content",
    colour = "Treatment",
    shape = "Cell line"
  )
ggarrange(
  a, b, ncol = 2, nrow = 1, 
  common.legend = TRUE, legend = "bottom"
) %>%
  annotate_figure("StrippedSerum Dataset (H. sapiens)")

Comparison of residual bias potentially introduced by incomplete rRNA removal. Regression lines are shown along with standard error bands for each comparison.

PCA

genes2keep <- dgeList %>%
  cpm() %>%
  is_greater_than(1) %>%
  rowSums() %>%
  is_weakly_greater_than(6)
dgeFilt <- dgeList[genes2keep,, keep.lib.sizes = FALSE] %>%
  calcNormFactors()

pca <- cpm(dgeFilt, log = TRUE) %>%
  t() %>%
  prcomp()

pcaCor <- pca$x %>%
  as.data.frame() %>%
  rownames_to_column("sample") %>%
  left_join(sumGc) %>%
  as_tibble() %>% 
  left_join(dgeList$samples) %>%
  dplyr::select(
    PC1, PC2, PC3, 
    Mean_GC = mn_gc, 
    Mean_Length = mn_logLen, 
    rRna_Proportion = rRNA
  ) %>% 
  cor()

a <- pca$x %>%
  as.data.frame() %>%
  rownames_to_column("sample") %>%
  left_join(dgeList$samples) %>%
  as_tibble() %>%
  ggplot(aes(PC1, PC2)) +
  geom_point(aes(colour = treat, shape = cell_line), size = 2) +
  labs(
    x = paste0("PC1 (", percent(summary(pca)$importance["Proportion of Variance","PC1"]),")"),
    y = paste0("PC2 (", percent(summary(pca)$importance["Proportion of Variance","PC2"]),")"),
    colour = "Treatment",
    shape = "Cell line"
  )
b <- pca$x %>%
  as.data.frame() %>%
  rownames_to_column("sample") %>%
  left_join(dgeList$samples) %>%
  ggplot(aes(PC1, rRNA, label = rRNA)) +
  geom_point(aes(colour = treat, shape = cell_line), size = 2) +
  geom_smooth(method = "lm") +
  geom_text_repel(show.legend = FALSE) +
  scale_y_continuous(labels = percent) +
  labs(
    x = paste0("PC1 (", percent(summary(pca)$importance["Proportion of Variance","PC1"]),")"),
    y = "rRNA Proportion of Initial Library",
    colour = "Treatment",
    shape = "Cell line"
  )
c <- pca$x %>%
  as.data.frame() %>%
  rownames_to_column("sample") %>%
  left_join(sumGc) %>%
  left_join(dgeList$samples) %>%
  as_tibble() %>%
  ggplot(aes(PC1, mn_logLen)) +
  geom_point(aes(colour = treat, shape = cell_line), size = 2) +
  geom_smooth(method = "lm") +
  labs(
    x = paste0("PC1 (", percent(summary(pca)$importance["Proportion of Variance","PC1"]),")"),
    y = "Mean log(Length)",
    colour = "Treatment",
    shape = "Cell line"
  )
d <- pca$x %>%
  as.data.frame() %>%
  rownames_to_column("sample") %>%
  left_join(sumGc) %>%
  left_join(dgeList$samples) %>%
  as_tibble() %>%
  ggplot(aes(PC1, mn_gc)) +
  geom_point(aes(colour = treat, shape = cell_line), size = 2) +
  geom_smooth(method = "lm") +
  scale_y_continuous(labels = percent) +
  labs(
    x = paste0("PC1 (", percent(summary(pca)$importance["Proportion of Variance","PC1"]),")"),
    y = "Mean GC",
    colour = "Treatment",
    shape = "Cell line"
  )
ggarrange(
  a, b, c, d, ncol = 2, nrow = 2,
  common.legend = TRUE, legend = "bottom"
) %>%
  annotate_figure("StrippedSerum")

PCA plot showing rRNA proportion, mean GC content and mean log(length) after summarisation to gene-level.

corrplot(
  pcaCor,
  type = "lower", 
  diag = FALSE, 
  addCoef.col = 1, addCoefasPercent = TRUE
)

Correlations between the first three principal components and measured variables: mean GC content, mean log(length) and rRNA proportion.

Differential expression

design <- model.matrix(~rRNA, data = dgeFilt$samples)

voom <- voom(dgeFilt, design = design)

fit <- lmFit(voom, design = design)

eBayes <- eBayes(fit)

topTable <- eBayes %>%
  topTable(coef = colnames(design)[2], sort.by = "p", n = Inf) %>%
  set_colnames(str_remove(colnames(.), "ID\\.")) %>%
  mutate(Bonf = p.adjust(P.Value, "bonferroni")) %>%
  mutate(DE = Bonf < 0.05) %>%
  unite(Location, c(seqnames, start, end, width, strand), sep = ":") %>%
  dplyr::select(
    Geneid = gene_id,
    Symbol = gene_name,
    AveExpr,
    logFC,
    P.Value,
    FDR = adj.P.Val,
    Location,
    t,
    DE,
    everything(),
    -B
  ) %>%
  as_tibble()

topTable %>% 
  dplyr::select(Geneid, Symbol, AveExpr, logFC, P.Value, FDR, DE) %>%
  mutate(
    AveExpr = format(round(AveExpr, 2), nsmall = 2),
    logFC = format(round(logFC, 2), nsmall = 2),
    P.Value = sprintf("%.2e", P.Value),
    FDR = sprintf("%.2e", FDR)
  ) %>%
  dplyr::slice(1:200) %>%
  datatable(
    options = list(pageLength = 20), 
    class = "striped hover condensed responsive", 
    filter = "top",
    caption = paste(
      "The top 100 differentially expressed genes.",
      nrow(dplyr::filter(topTable, DE)),
      "of",
      nrow(topTable),
      "genes were classified as DE with an Bonferroni p-value < 0.05."
    )
  )

k-mer analysis

jellyfish v2.3.0 was used to count kmers of trimmed fastq files that had been filtered for rRNA sequences. This was performed for 5 values: \(k = 5, 6, 7, 8, 9, 10\). Lower values of \(k\) lose specificity in comparison to higher values, however as \(k\) increases, the exponential increase of possible kmers causes limitations due to computational processing time.

Setup

k = 5

Counts

k5files <- list.files("/hpcfs/users/a1647910/20200310_rRNADepletion/4_T47D_ZR75_DHT_StrippedSerum/6_jellyfish2pass/k5", pattern = "_dumps.txt", full.names = TRUE)
k5counts <- lapply(k5files, function(x){
  read_delim(x, col_names = c("mer", basename(x)), delim = " ") %>%
    set_colnames(str_remove_all(colnames(.), "_dumps\\.txt"))
}) %>%
  purrr::reduce(full_join) %>%
  dplyr::select(mer, contains(c("D", "V")))

k5dge <- k5counts %>%
  as.data.frame() %>%
  column_to_rownames("mer") %>%
  DGEList() %>%
  calcNormFactors()
k5dge$samples %<>%
  rownames_to_column("rowname") %>%
  mutate(sample = rowname) %>%
  left_join(addInfo) %>%
  column_to_rownames("rowname")

Properties

k5dist <- k5dge %>%
  cpm(log = TRUE) %>%
  as.data.frame() %>%
  pivot_longer(everything(), names_to = "sample", values_to = "count") %>%
  ggplot(aes(x=count, colour = sample)) +
  geom_density() +
  labs(x = "intensity", title = "Distribution of 5-mers")

k5labels <- k5dge$samples %>% 
  mutate(label = paste0(sample, "\n", percent(rRNA, accuracy = 0.01), " rRNA")) %>% 
  .$label
k5heat <- k5dge %>% 
  cpm(log = TRUE) %>%
  as.data.frame() %>%
  t() %>%
  pheatmap(silent = TRUE, cluster_cols = FALSE,
           show_colnames = FALSE, fontsize = 9,
           fontsize_row = 10, border_color = NA,
           main = "5-mer counts heatmap", labels_row = k5labels)
k5heat$tree_row$labels <- k5labels
k5den <- ggdendrogram(k5heat$tree_row, rotate = TRUE) +
  labs(title = "Hierarchical clustering of 5-mer counts") +
  theme(plot.title = element_text(size = 12))

# Assess cpm values to make sure PCA results are not heavily skewed by highly expressed genes
k5pca <- k5dge %>%
  cpm(log = TRUE) %>%
  t() %>%
  prcomp()

# Quick inspection to check whether first two PCA components capture most of the variability
summary(k5pca)$importance %>% pander(split.tables = Inf)

k5pcaPlot <- k5pca$x %>%
  as.data.frame() %>%
  rownames_to_column("sample") %>%
  as_tibble() %>%
  dplyr::select(sample, PC1, PC2) %>%
  left_join(k5dge$samples) %>%
  ggplot(aes(PC1, PC2, colour = treat, shape = cell_line, label = rRNA)) +
  geom_point(alpha = 0.8, size = 3) +
  geom_text_repel(show.legend = FALSE) +
  labs(
    x = paste0("PC1 (", percent(summary(k5pca)$importance[2, "PC1"]), ")"),
    y = paste0("PC2 (", percent(summary(k5pca)$importance[2, "PC2"]), ")"),
    colour = "Treatment",
    shape = "Cell line",
    title = "k = 5"
  )

k5pcaRrna <- k5pca$x %>%
  as.data.frame() %>%
  rownames_to_column("sample") %>%
  as_tibble() %>%
  dplyr::select(sample, PC1, PC2) %>%
  left_join(k5dge$samples) %>%
  ggplot(aes(PC1, rRNA, label = rRNA)) +
  geom_point(aes(colour = treat, shape = cell_line), alpha = 0.8, size = 3) +
  geom_text_repel(show.legend = FALSE) +
  geom_smooth(method = "lm") +
  labs(
    x = paste0("PC1 (", percent(summary(k5pca)$importance[2, "PC1"]), ")"),
    y = "rRNA proportion",
    colour = "Treatment",
    shape = "Cell line",
    title = "k = 5"
  ) +
  scale_y_continuous(labels = percent)

Differential expression

k5design <- model.matrix(~rRNA, data = k5dge$samples)
k5voom <- voom(k5dge, design = k5design)
k5fit <- lmFit(k5voom, design = k5design)
k5eBayes <- eBayes(k5fit)

k5topTable <- k5eBayes %>%
  topTable(coef = colnames(k5design)[2], sort.by = "p", n = Inf) %>%
  set_colnames(str_remove(colnames(.), "ID\\.")) %>%
  rownames_to_column("mer") %>%
  mutate(BY = p.adjust(P.Value, "BY")) %>%
  mutate(DE = BY < 0.05) %>%
  dplyr::select(
    mer,
    AveExpr,
    logFC,
    P.Value,
    FDR = adj.P.Val,
    BY,
    t,
    DE,
    everything(),
    -B
  ) %>%
  as_tibble()

k = 6

Counts

k6files <- list.files("/hpcfs/users/a1647910/20200310_rRNADepletion/4_T47D_ZR75_DHT_StrippedSerum/6_jellyfish2pass/k6", pattern = "_dumps.txt", full.names = TRUE)
k6counts <- lapply(k6files, function(x){
  read_delim(x, col_names = c("mer", basename(x)), delim = " ") %>%
    set_colnames(str_remove_all(colnames(.), "_dumps\\.txt"))
}) %>%
  purrr::reduce(full_join) %>%
  dplyr::select(mer, contains(c("D", "V")))

k6dge <- k6counts %>%
  as.data.frame() %>%
  column_to_rownames("mer") %>%
  DGEList() %>%
  calcNormFactors()
k6dge$samples %<>%
  rownames_to_column("rowname") %>%
  mutate(sample = rowname) %>%
  left_join(addInfo) %>%
  column_to_rownames("rowname")
k6dge$samples$group <- colnames(k6dge) %>%
  str_extract("(D|V)") %>%
  factor(levels = c("D", "V"))

Properties

k6dist <- k6dge %>%
  cpm(log = TRUE) %>%
  as.data.frame() %>%
  pivot_longer(everything(), names_to = "sample", values_to = "count") %>%
  ggplot(aes(x=count, colour = sample)) +
  geom_density() +
  labs(x = "intensity", title = "Distribution of 6-mers")

k6labels <- k6dge$samples %>% 
  mutate(label = paste0(sample, "\n", percent(rRNA, accuracy = 0.01), " rRNA")) %>% 
  .$label
k6heat <- k6dge %>% 
  cpm(log = TRUE) %>%
  as.data.frame() %>%
  t() %>%
  pheatmap(silent = TRUE, cluster_cols = FALSE,
           show_colnames = FALSE, fontsize = 9,
           fontsize_row = 10, border_color = NA,
           main = "6-mer counts heatmap", labels_row = k6labels)
k6heat$tree_row$labels <- k6labels
k6den <- ggdendrogram(k6heat$tree_row, rotate = TRUE) +
  labs(title = "Hierarchical clustering of 6-mer counts") +
  theme(plot.title = element_text(size = 12))

# Assess cpm values to make sure PCA results are not heavily skewed by highly expressed genes
k6pca <- k6dge %>%
  cpm(log = TRUE) %>%
  t() %>%
  prcomp()

# Quick inspection to check whether first two PCA components capture most of the variability
summary(k6pca)$importance %>% pander(split.tables = Inf)

# Plot PCA
k6pcaPlot <- k6pca$x %>%
  as.data.frame() %>%
  rownames_to_column("sample") %>%
  as_tibble() %>%
  dplyr::select(sample, PC1, PC2) %>%
  left_join(k6dge$samples) %>%
  ggplot(aes(PC1, PC2, colour = treat, shape = cell_line, label = rRNA)) +
  geom_point(alpha = 0.8, size = 3) +
  geom_text_repel(show.legend = FALSE) +
  labs(
    x = paste0("PC1 (", percent(summary(k6pca)$importance[2, "PC1"]), ")"),
    y = paste0("PC2 (", percent(summary(k6pca)$importance[2, "PC2"]), ")"),
    colour = "Treatment",
    shape = "Cell line",
    title = "k = 6"
  )

k6pcaRrna <- k6pca$x %>%
  as.data.frame() %>%
  rownames_to_column("sample") %>%
  as_tibble() %>%
  dplyr::select(sample, PC1, PC2) %>%
  left_join(k6dge$samples) %>%
  ggplot(aes(PC1, rRNA, label = rRNA)) +
  geom_point(aes(colour = treat, shape = cell_line), alpha = 0.8, size = 3) +
  geom_text_repel(show.legend = FALSE) +
  geom_smooth(method = "lm") +
  labs(
    x = paste0("PC1 (", percent(summary(k6pca)$importance[2, "PC1"]), ")"),
    y = "rRNA proportion",
    colour = "Treatment",
    shape = "Cell line",
    title = "k = 6"
  ) +
  scale_y_continuous(labels = percent)

Differential expression

k6design <- model.matrix(~rRNA, data = k6dge$samples)
k6voom <- voom(k6dge, design = k6design)
k6fit <- lmFit(k6voom, design = k6design)
k6eBayes <- eBayes(k6fit)

k6topTable <- k6eBayes %>%
  topTable(coef = colnames(k6design)[2], sort.by = "p", n = Inf) %>%
  set_colnames(str_remove(colnames(.), "ID\\.")) %>%
  rownames_to_column("mer") %>%
  mutate(BY = p.adjust(P.Value, "BY")) %>%
  mutate(DE = BY < 0.05) %>%
  dplyr::select(
    mer,
    AveExpr,
    logFC,
    P.Value,
    FDR = adj.P.Val,
    BY,
    t,
    DE,
    everything(),
    -B
  ) %>%
  as_tibble()

k = 7

Counts

k7files <- list.files("/hpcfs/users/a1647910/20200310_rRNADepletion/4_T47D_ZR75_DHT_StrippedSerum/6_jellyfish2pass/k7", pattern = "_dumps.txt", full.names = TRUE)
k7counts <- lapply(k7files, function(x){
  read_delim(x, col_names = c("mer", basename(x)), delim = " ") %>%
    set_colnames(str_remove_all(colnames(.), "_dumps\\.txt"))
}) %>%
  purrr::reduce(full_join) %>%
  dplyr::select(mer, contains(c("D", "V")))

k7dge <- k7counts %>%
  as.data.frame() %>%
  column_to_rownames("mer") %>%
  DGEList() %>%
  calcNormFactors()
k7dge$samples %<>%
  rownames_to_column("rowname") %>%
  mutate(sample = rowname) %>%
  left_join(addInfo) %>%
  column_to_rownames("rowname")
k7dge$samples$group <- colnames(k7dge) %>%
  str_extract("(D|V)") %>%
  factor(levels = c("D", "V"))

Properties

k7dist <- k7dge %>%
  cpm(log = TRUE) %>%
  as.data.frame() %>%
  pivot_longer(everything(), names_to = "sample", values_to = "count") %>%
  ggplot(aes(x=count, colour = sample)) +
  geom_density() +
  labs(x = "intensity", title = "Distribution of 7-mers")

k7labels <- k7dge$samples %>% 
  mutate(label = paste0(sample, "\n", percent(rRNA, accuracy = 0.01), " rRNA")) %>% 
  .$label
k7heat <- k7dge %>% 
  cpm(log = TRUE) %>%
  as.data.frame() %>%
  t() %>%
  pheatmap(silent = TRUE, cluster_cols = FALSE,
           show_colnames = FALSE, fontsize = 9,
           fontsize_row = 10, border_color = NA,
           main = "7-mer counts heatmap", labels_row = k7labels)
k7heat$tree_row$labels <- k7labels
k7den <- ggdendrogram(k7heat$tree_row, rotate = TRUE) +
  labs(title = "Hierarchical clustering of 7-mer counts") +
  theme(plot.title = element_text(size = 12))

# Assess cpm values to make sure PCA results are not heavily skewed by highly expressed genes
k7pca <- k7dge %>%
  cpm(log = TRUE) %>%
  t() %>%
  prcomp()

# Quick inspection to check whether first two PCA components capture most of the variability
summary(k7pca)$importance %>% pander(split.tables = Inf)

# Plot PCA
k7pcaPlot <- k7pca$x %>%
  as.data.frame() %>%
  rownames_to_column("sample") %>%
  as_tibble() %>%
  dplyr::select(sample, PC1, PC2) %>%
  left_join(k7dge$samples) %>%
  ggplot(aes(PC1, PC2, colour = treat, shape = cell_line, label = rRNA)) +
  geom_point(alpha = 0.8, size = 3) +
  geom_text_repel(show.legend = FALSE) +
  labs(
    x = paste0("PC1 (", percent(summary(k7pca)$importance[2, "PC1"]), ")"),
    y = paste0("PC2 (", percent(summary(k7pca)$importance[2, "PC2"]), ")"),
    colour = "Treatment",
    shape = "Cell line",
    title = "k = 7"
  )

k7pcaRrna <- k7pca$x %>%
  as.data.frame() %>%
  rownames_to_column("sample") %>%
  as_tibble() %>%
  dplyr::select(sample, PC1, PC2) %>%
  left_join(k7dge$samples) %>%
  ggplot(aes(PC1, rRNA, label = rRNA)) +
  geom_point(aes(colour = treat, shape = cell_line), alpha = 0.8, size = 3) +
  geom_text_repel(show.legend = FALSE) +
  geom_smooth(method = "lm") +
  labs(
    x = paste0("PC1 (", percent(summary(k7pca)$importance[2, "PC1"]), ")"),
    y = "rRNA proportion",
    colour = "Treatment",
    title = "k = 7"
  ) +
  scale_y_continuous(labels = percent)

Differential expression

k7design <- model.matrix(~rRNA, data = k7dge$samples)
k7voom <- voom(k7dge, design = k7design)
k7fit <- lmFit(k7voom, design = k7design)
k7eBayes <- eBayes(k7fit)

k7topTable <- k7eBayes %>%
  topTable(coef = colnames(k7design)[2], sort.by = "p", n = Inf) %>%
  set_colnames(str_remove(colnames(.), "ID\\.")) %>%
  rownames_to_column("mer") %>%
  mutate(BY = p.adjust(P.Value, "BY")) %>%
  mutate(DE = BY < 0.05) %>%
  dplyr::select(
    mer,
    AveExpr,
    logFC,
    P.Value,
    FDR = adj.P.Val,
    BY,
    t,
    DE,
    everything(),
    -B
  ) %>%
  as_tibble()

k = 8

Counts

k8files <- list.files("/hpcfs/users/a1647910/20200310_rRNADepletion/4_T47D_ZR75_DHT_StrippedSerum/6_jellyfish2pass/k8", pattern = "_dumps.txt", full.names = TRUE)
k8counts <- lapply(k8files, function(x){
  read_delim(x, col_names = c("mer", basename(x)), delim = " ") %>%
    set_colnames(str_remove_all(colnames(.), "_dumps\\.txt"))
}) %>%
  purrr::reduce(full_join) %>%
  dplyr::select(mer, contains(c("D", "V")))

k8dge <- k8counts %>%
  as.data.frame() %>%
  column_to_rownames("mer") %>%
  DGEList() %>%
  calcNormFactors()
k8dge$samples %<>%
  rownames_to_column("rowname") %>%
  mutate(sample = rowname) %>%
  left_join(addInfo) %>%
  column_to_rownames("rowname")
k8dge$samples$group <- colnames(k8dge) %>%
  str_extract("(D|V)") %>%
  factor(levels = c("D", "V"))

Properties

k8dist <- k8dge %>%
  cpm(log = TRUE) %>%
  as.data.frame() %>%
  pivot_longer(everything(), names_to = "sample", values_to = "count") %>%
  ggplot(aes(x=count, colour = sample)) +
  geom_density() +
  labs(x = "intensity", title = "Distribution of 8-mers")

k8labels <- k8dge$samples %>% 
  mutate(label = paste0(sample, "\n", percent(rRNA, accuracy = 0.01), " rRNA")) %>% 
  .$label
k8heat <- k8dge %>% 
  cpm(log = TRUE) %>%
  as.data.frame() %>%
  t() %>%
  pheatmap(silent = TRUE, cluster_cols = FALSE,
           show_colnames = FALSE, fontsize = 9,
           fontsize_row = 10, border_color = NA,
           main = "8-mer counts heatmap", labels_row = k8labels)
k8heat$tree_row$labels <- k8labels
k8den <- ggdendrogram(k8heat$tree_row, rotate = TRUE) +
  labs(title = "Hierarchical clustering of 8-mer counts") +
  theme(plot.title = element_text(size = 12))

# Assess cpm values to make sure PCA results are not heavily skewed by highly expressed genes
k8pca <- k8dge %>%
  cpm(log = TRUE) %>%
  t() %>%
  prcomp()

# Quick inspection to check whether first two PCA components capture most of the variability
summary(k8pca)$importance %>% pander(split.tables = Inf)

# Plot PCA
k8pcaPlot <- k8pca$x %>%
  as.data.frame() %>%
  rownames_to_column("sample") %>%
  as_tibble() %>%
  dplyr::select(sample, PC1, PC2) %>%
  left_join(k8dge$samples) %>%
  ggplot(aes(PC1, PC2, colour = treat, shape = cell_line, label = rRNA)) +
  geom_point(alpha = 0.8, size = 3) +
  geom_text_repel(show.legend = FALSE) +
  labs(
    x = paste0("PC1 (", percent(summary(k8pca)$importance[2, "PC1"]), ")"),
    y = paste0("PC2 (", percent(summary(k8pca)$importance[2, "PC2"]), ")"),
    colour = "Treatment",
    shape = "Cell line",
    title = "k = 8"
  )

k8pcaRrna <- k8pca$x %>%
  as.data.frame() %>%
  rownames_to_column("sample") %>%
  as_tibble() %>%
  dplyr::select(sample, PC1, PC2) %>%
  left_join(k8dge$samples) %>%
  ggplot(aes(PC1, rRNA, label = rRNA)) +
  geom_point(aes(colour = treat, shape = cell_line), alpha = 0.8, size = 3) +
  geom_text_repel(show.legend = FALSE) +
  geom_smooth(method = "lm") +
  labs(
    x = paste0("PC1 (", percent(summary(k8pca)$importance[2, "PC1"]), ")"),
    y = "rRNA proportion",
    colour = "Treatment",
    shape = "Cell line",
    title = "k = 8"
  ) +
  scale_y_continuous(labels = percent)

Differential expression

k8design <- model.matrix(~rRNA, data = k8dge$samples)
k8voom <- voom(k8dge, design = k8design)
k8fit <- lmFit(k8voom, design = k8design)
k8eBayes <- eBayes(k8fit)

k8topTable <- k8eBayes %>%
  topTable(coef = colnames(k8design)[2], sort.by = "p", n = Inf) %>%
  set_colnames(str_remove(colnames(.), "ID\\.")) %>%
  rownames_to_column("mer") %>%
  mutate(BY = p.adjust(P.Value, "BY")) %>%
  mutate(DE = BY < 0.05) %>%
  dplyr::select(
    mer,
    AveExpr,
    logFC,
    P.Value,
    FDR = adj.P.Val,
    BY,
    t,
    DE,
    everything(),
    -B
  ) %>%
  as_tibble()

k = 9

Counts

k9files <- list.files("/hpcfs/users/a1647910/20200310_rRNADepletion/4_T47D_ZR75_DHT_StrippedSerum/6_jellyfish2pass/k9", pattern = "_dumps.txt", full.names = TRUE)
k9counts <- lapply(k9files, function(x){
  read_delim(x, col_names = c("mer", basename(x)), delim = " ") %>%
    set_colnames(str_remove_all(colnames(.), "_dumps\\.txt"))
}) %>%
  purrr::reduce(full_join) %>%
  dplyr::select(mer, contains(c("D", "V")))

k9dge <- k9counts %>%
  as.data.frame() %>%
  column_to_rownames("mer") %>%
  DGEList() %>%
  calcNormFactors()
k9dge$samples %<>%
  rownames_to_column("rowname") %>%
  mutate(sample = rowname) %>%
  left_join(addInfo) %>%
  column_to_rownames("rowname")
k9dge$samples$group <- colnames(k9dge) %>%
  str_extract("(D|V)") %>%
  factor(levels = c("D", "V"))

Properties

k9dist <- k9dge %>%
  cpm(log = TRUE) %>%
  as.data.frame() %>%
  pivot_longer(everything(), names_to = "sample", values_to = "count") %>%
  ggplot(aes(x=count, colour = sample)) +
  geom_density() +
  labs(x = "intensity", title = "Distribution of 9-mers")

k9labels <- k9dge$samples %>% 
  mutate(label = paste0(sample, "\n", percent(rRNA, accuracy = 0.01), " rRNA")) %>% 
  .$label
k9heat <- k9dge %>% 
  cpm(log = TRUE) %>%
  as.data.frame() %>%
  t() %>%
  pheatmap(silent = TRUE, cluster_cols = FALSE,
           show_colnames = FALSE, fontsize = 9,
           fontsize_row = 10, border_color = NA,
           main = "9-mer counts heatmap", labels_row = k9labels)
k9heat$tree_row$labels <- k9labels
k9den <- ggdendrogram(k9heat$tree_row, rotate = TRUE) +
  labs(title = "Hierarchical clustering of 9-mer counts") +
  theme(plot.title = element_text(size = 12))

# Assess cpm values to make sure PCA results are not heavily skewed by highly expressed genes
k9pca <- k9dge %>%
  cpm(log = TRUE) %>%
  t() %>%
  prcomp()

# Quick inspection to check whether first two PCA components capture most of the variability
summary(k9pca)$importance %>% pander(split.tables = Inf)

# Plot PCA
k9pcaPlot <- k9pca$x %>%
  as.data.frame() %>%
  rownames_to_column("sample") %>%
  as_tibble() %>%
  dplyr::select(sample, PC1, PC2) %>%
  left_join(k9dge$samples) %>%
  ggplot(aes(PC1, PC2, colour = treat, shape = cell_line, label = rRNA)) +
  geom_point(alpha = 0.8, size = 3) +
  geom_text_repel(show.legend = FALSE) +
  labs(
    x = paste0("PC1 (", percent(summary(k9pca)$importance[2, "PC1"]), ")"),
    y = paste0("PC2 (", percent(summary(k9pca)$importance[2, "PC2"]), ")"),
    colour = "Treatment",
    shape = "Cell line",
    title = "k = 9"
  )

k9pcaRrna <- k9pca$x %>%
  as.data.frame() %>%
  rownames_to_column("sample") %>%
  as_tibble() %>%
  dplyr::select(sample, PC1, PC2) %>%
  left_join(k9dge$samples) %>%
  ggplot(aes(PC1, rRNA, label = rRNA)) +
  geom_point(aes(colour = treat, shape = cell_line), alpha = 0.8, size = 3) +
  geom_text_repel(show.legend = FALSE) +
  geom_smooth(method = "lm") +
  labs(
    x = paste0("PC1 (", percent(summary(k9pca)$importance[2, "PC1"]), ")"),
    y = "rRNA proportion",
    colour = "Treatment",
    shape = "Cell line",
    title = "k = 9"
  ) +
  scale_y_continuous(labels = percent)

Differential expression

k9design <- model.matrix(~rRNA, data = k9dge$samples)
k9voom <- voom(k9dge, design = k9design)
k9fit <- lmFit(k9voom, design = k9design)
k9eBayes <- eBayes(k9fit)

k9topTable <- k9eBayes %>%
  topTable(coef = colnames(k9design)[2], sort.by = "p", n = Inf) %>%
  set_colnames(str_remove(colnames(.), "ID\\.")) %>%
  rownames_to_column("mer") %>%
  mutate(BY = p.adjust(P.Value, "BY")) %>%
  mutate(DE = BY < 0.05) %>%
  dplyr::select(
    mer,
    AveExpr,
    logFC,
    P.Value,
    FDR = adj.P.Val,
    BY,
    t,
    DE,
    everything(),
    -B
  ) %>%
  as_tibble()

k = 10

Counts

k10files <- list.files("/hpcfs/users/a1647910/20200310_rRNADepletion/4_T47D_ZR75_DHT_StrippedSerum/6_jellyfish2pass/k10", pattern = "_dumps.txt", full.names = TRUE)
k10counts <- lapply(k10files, function(x){
  read_delim(x, col_names = c("mer", basename(x)), delim = " ") %>%
    set_colnames(str_remove_all(colnames(.), "_dumps\\.txt"))
}) %>%
  purrr::reduce(full_join) %>%
  dplyr::select(mer, contains(c("D", "V")))
k10counts[is.na(k10counts)] <- 0

k10dge <- k10counts %>%
  as.data.frame() %>%
  column_to_rownames("mer") %>%
  DGEList() %>%
  calcNormFactors()
k10dge$samples %<>%
  rownames_to_column("rowname") %>%
  mutate(sample = rowname) %>%
  left_join(addInfo) %>%
  column_to_rownames("rowname")
k10dge$samples$group <- colnames(k10dge) %>%
  str_extract("(D|V)") %>%
  factor(levels = c("D", "V"))

Properties

k10dist <- k10dge %>%
  cpm(log = TRUE) %>%
  as.data.frame() %>%
  pivot_longer(everything(), names_to = "sample", values_to = "count") %>%
  ggplot(aes(x=count, colour = sample)) +
  geom_density() +
  labs(x = "intensity", title = "Distribution of 10-mers")

k10labels <- k10dge$samples %>% 
  mutate(label = paste0(sample, "\n", percent(rRNA, accuracy = 0.01), " rRNA")) %>% 
  .$label
k10heat <- k10dge %>% 
  cpm(log = TRUE) %>%
  as.data.frame() %>%
  t() %>%
  pheatmap(silent = TRUE, cluster_cols = FALSE,
           show_colnames = FALSE, fontsize = 9,
           fontsize_row = 10, border_color = NA,
           main = "10-mer counts heatmap", labels_row = k10labels)
k10heat$tree_row$labels <- k10labels
k10den <- ggdendrogram(k10heat$tree_row, rotate = TRUE) +
  labs(title = "Hierarchical clustering of 10-mer counts") +
  theme(plot.title = element_text(size = 12))

# Assess cpm values to make sure PCA results are not heavily skewed by highly expressed genes
k10pca <- k10dge %>%
  cpm(log = TRUE) %>%
  t() %>%
  prcomp()

# Quick inspection to check whether first two PCA components capture most of the variability
summary(k10pca)$importance %>% pander(split.tables = Inf)

# Plot PCA
k10pcaPlot <- k10pca$x %>%
  as.data.frame() %>%
  rownames_to_column("sample") %>%
  as_tibble() %>%
  dplyr::select(sample, PC1, PC2) %>%
  left_join(k10dge$samples) %>%
  ggplot(aes(PC1, PC2, colour = treat, shape = cell_line, label = rRNA)) +
  geom_point(alpha = 0.8, size = 3) +
  geom_text_repel(show.legend = FALSE) +
  labs(
    x = paste0("PC1 (", percent(summary(k10pca)$importance[2, "PC1"]), ")"),
    y = paste0("PC2 (", percent(summary(k10pca)$importance[2, "PC2"]), ")"),
    colour = "Treatment",
    shape = "Cell line",
    title = "k = 10"
  )

k10pcaRrna <- k10pca$x %>%
  as.data.frame() %>%
  rownames_to_column("sample") %>%
  as_tibble() %>%
  dplyr::select(sample, PC1, PC2) %>%
  left_join(k10dge$samples) %>%
  ggplot(aes(PC1, rRNA, label = rRNA)) +
  geom_point(aes(colour = treat, shape = cell_line), alpha = 0.8, size = 3) +
  geom_text_repel(show.legend = FALSE) +
  geom_smooth(method = "lm") +
  labs(
    x = paste0("PC1 (", percent(summary(k10pca)$importance[2, "PC1"]), ")"),
    y = "rRNA proportion",
    colour = "Treatment",
    shape = "Cell line",
    title = "k = 10"
  ) +
  scale_y_continuous(labels = percent)

Differential expression

k10design <- model.matrix(~rRNA, data = k10dge$samples)
k10voom <- voom(k10dge, design = k10design)
k10fit <- lmFit(k10voom, design = k10design)
k10eBayes <- eBayes(k10fit)

k10topTable <- k10eBayes %>%
  topTable(coef = colnames(k10design)[2], sort.by = "p", n = Inf) %>%
  set_colnames(str_remove(colnames(.), "ID\\.")) %>%
  rownames_to_column("mer") %>%
  mutate(BY = p.adjust(P.Value, "BY")) %>%
  mutate(DE = BY < 0.05) %>%
  dplyr::select(
    mer,
    AveExpr,
    logFC,
    P.Value,
    FDR = adj.P.Val,
    BY,
    t,
    DE,
    everything(),
    -B
  ) %>%
  as_tibble()

Distributions

ggarrange(
  k5dist, k6dist, k7dist, k8dist, k9dist, k10dist,
  ncol = 2, nrow = 3, common.legend = TRUE, legend = "bottom" 
)

Distributions of kmer counts for each value of \(k\)

PCA

ggarrange(
  k5pcaPlot, k6pcaPlot, k7pcaPlot, k8pcaPlot, k9pcaPlot, k10pcaPlot,
  ncol = 2, nrow = 3, common.legend = TRUE, legend = "bottom"
)

Principal component analysis for all values of \(k\).

ggarrange(
  k5pcaRrna, k6pcaRrna, k7pcaRrna, k8pcaRrna, k9pcaRrna, k10pcaRrna,
  ncol = 2, nrow = 3, common.legend = TRUE, legend = "bottom"
)

Principal component analysis for all values of \(k\).

Clustering

ggarrange(
  k5den, k6den, k7den, k8den, k9den, k10den,
  ncol = 2, nrow = 3, common.legend = TRUE, legend = "bottom"
)

Hierarchical clustering of samples based on kmer counts.

Differential expression

topRes <- function(x, cap){
  x %>% 
    dplyr::select(mer, AveExpr, logFC, P.Value, FDR, BY, DE) %>%
    mutate(
      AveExpr = format(round(AveExpr, 2), nsmall = 2),
      logFC = format(round(logFC, 2), nsmall = 2),
      P.Value = sprintf("%.2e", P.Value),
      FDR = sprintf("%.2e", FDR),
      BY = sprintf("%.2e", BY)
    ) %>%
    dplyr::slice(1:100) %>%
    datatable(
      options = list(pageLength = 10),
      class = "striped hover condensed responsive",
      filter = "top",
      caption = cap
    )
}

k = 5

topRes(k5topTable,
       cap = paste(
         "The top 100 differentially expressed 5-mers.",
         nrow(dplyr::filter(k5topTable, DE)),
         "of",
         nrow(k5topTable),
         "detected sequences were classified as DE with BY p-value < 0.05."
       )
)

k5topTable %>%
  ggplot(aes(P.Value)) +
  geom_histogram(
    binwidth = 0.02,
    colour = "black", fill = "grey90"
  ) +
  labs(title = "k = 5")

Histogram of p-values. Values follow the expected distribution when there are many differences.

k5topTable %>%
  ggplot(aes(logFC, -log10(P.Value), colour = DE)) +
  geom_point(alpha = 0.5) +
  scale_colour_manual(values = c("black", "grey50", "red")) +
  # geom_text_repel(
  #   data = . %>%
  #     # dplyr::filter(-log10(P.Value) > 4 | logFC > 4 | logFC < -2),
  #     dplyr::filter(-log10(P.Value) > 4 | logFC > 3 | logFC < -2.5),
  #   aes(label = mer, color = "black")
  # ) +
  labs(x = "logFC", y = expression(paste(-log[10], "(p)")), title = "k = 5") +
  theme_bw() +
  theme(legend.position = "none")

Volcano plot showing -log10(p-value) against logFC. Kmers classified as DE with BY p-value < 0.05 are highlighted in red.

---

k = 6

topRes(k6topTable,
       cap = paste(
         "The top 100 differentially expressed 6-mers.",
         nrow(dplyr::filter(k6topTable, DE)),
         "of",
         nrow(k6topTable),
         "detected sequences were classified as DE with a BY p-value < 0.05."
       )
)

k6topTable %>%
  ggplot(aes(P.Value)) +
  geom_histogram(
    binwidth = 0.02,
    colour = "black", fill = "grey90"
  ) +
  labs(title = "k = 6")

Histogram of p-values. Values follow the expected distribution when there are some differences.

k6topTable %>%
  ggplot(aes(logFC, -log10(P.Value), colour = DE)) +
  geom_point(alpha = 0.5) +
  scale_colour_manual(values = c("black", "grey50", "red")) +
  # geom_text_repel(
  #   data = . %>%
  #     dplyr::filter(-log10(P.Value) > 6 | logFC > 6 | logFC < -2.3),
  #   aes(label = mer, color = "black")
  # ) +
  labs(x = "logFC", y = expression(paste(-log[10], "(p)")), title = "k = 6") +
  theme_bw() +
  theme(legend.position = "none")

Volcano plot showing -log10(p-value) against logFC. Kmers classified as DE with BY p-value < 0.05 are highlighted in red.

---

k = 7

topRes(k7topTable,
       cap = paste(
         "The top 100 differentially expressed 7-mers.",
         nrow(dplyr::filter(k7topTable, DE)),
         "of",
         nrow(k7topTable),
         "detected sequences were classified as DE with a BY p-value < 0.05."
       )
)

k7topTable %>%
  ggplot(aes(P.Value)) +
  geom_histogram(
    binwidth = 0.02,
    colour = "black", fill = "grey90"
  ) +
  labs(title = "k = 7")

Histogram of p-values. Values follow the expected distribution when there are many differences.

k7topTable %>%
  ggplot(aes(logFC, -log10(P.Value), colour = DE)) +
  geom_point(alpha = 0.5) +
  scale_colour_manual(values = c("black", "grey50", "red")) +
  # geom_text_repel(
  #   data = . %>%
  #     dplyr::filter(-log10(P.Value) > 6.4 | logFC > 10 | logFC < -5),
  #   aes(label = mer, color = "black")
  # ) +
  labs(x = "logFC", y = expression(paste(-log[10], "(p)")), title = "k = 7") +
  theme_bw() +
  theme(legend.position = "none")

Volcano plot showing -log10(p-value) against logFC. Kmers classified as DE with BY p-value < 0.05 are highlighted in red.

---

k = 8

topRes(k8topTable,
       cap = paste(
         "The top 100 differentially expressed 8-mers.",
         nrow(dplyr::filter(k8topTable, DE)),
         "of",
         nrow(k8topTable),
         "detected sequences were classified as DE with a BY p-value < 0.05."
       )
)

k8topTable %>%
  ggplot(aes(P.Value)) +
  geom_histogram(
    binwidth = 0.02,
    colour = "black", fill = "grey90"
  ) +
  labs(title = "k = 8")

Histogram of p-values. Values follow the expected distribution when there are many differences.

k8topTable %>%
  ggplot(aes(logFC, -log10(P.Value), colour = DE)) +
  geom_point(alpha = 0.5) +
  scale_colour_manual(values = c("black", "grey50", "red")) +
  # geom_text_repel(
  #   data = . %>%
  #     dplyr::filter(-log10(P.Value) > 7.2 | logFC > 12.5 | logFC < -8),
  #   aes(label = mer, color = "black")
  # ) +
  labs(x = "logFC", y = expression(paste(-log[10], "(p)")), title = "k = 8") +
  theme_bw() +
  theme(legend.position = "none")

Volcano plot showing -log10(p-value) against logFC. Kmers classified as DE with BY p-value < 0.05 are highlighted in red.

---

k = 9

topRes(k9topTable,
       cap = paste(
         "The top 100 differentially expressed 9-mers.",
         nrow(dplyr::filter(k9topTable, DE)),
         "of",
         nrow(k9topTable),
         "detected sequences were classified as DE with a BY p-value < 0.05."
       )
)

k9topTable %>%
  ggplot(aes(P.Value)) +
  geom_histogram(
    binwidth = 0.02,
    colour = "black", fill = "grey90"
  ) +
  labs(title = "k = 9")

Histogram of p-values. Values follow the expected distribution when there are many differences.

k9topTable %>%
  ggplot(aes(logFC, -log10(P.Value), colour = DE)) +
  geom_point(alpha = 0.5) +
  scale_colour_manual(values = c("black", "grey50", "red")) +
  # geom_text_repel(
  #   data = . %>%
  #     dplyr::filter(DE & -log10(P.Value) > 7.5 | logFC < -10 | logFC > 15),
  #   aes(label = mer, color = "black")
  # ) +
  labs(x = "logFC", y = expression(paste(-log[10], "(p)")), title = "k = 9") +
  theme_bw() +
  theme(legend.position = "none")

Volcano plot showing -log10(p-value) against logFC. Kmers classified as DE with BY p-value < 0.05 are highlighted in red.

---

k = 10

topRes(k10topTable,
       cap = paste(
         "The top 100 differentially expressed 10-mers.",
         nrow(dplyr::filter(k10topTable, DE)),
         "of",
         nrow(k10topTable),
         "detected sequences were classified as DE with a BY p-value < 0.05."
       )
)

k10topTable %>%
  ggplot(aes(P.Value)) +
  geom_histogram(
    binwidth = 0.02,
    colour = "black", fill = "grey90"
  ) +
  labs(title = "k = 10")

Histogram of p-values. Values follow the expected distribution when there are many differences.

k10topTable %>%
  ggplot(aes(logFC, -log10(P.Value), colour = DE)) +
  geom_point(alpha = 0.5) +
  scale_colour_manual(values = c("black", "grey50", "red")) +
  # geom_text_repel(
  #   data = . %>%
  #     dplyr::filter(-log10(P.Value) > 8.5 | logFC > 20 | logFC < -13.5),
  #   aes(label = mer, color = "black")
  # ) +
  labs(x = "logFC", y = expression(paste(-log[10], "(p)")), title = "k = 10") +
  theme_bw() +
  theme(legend.position = "none")

Volcano plot showing -log10(p-value) against logFC. Kmers classified as DE with BY p-value < 0.05 are highlighted in red.